Understanding lung cancer genomics in patients of African ancestry

PROJECT SUMMARY Lung cancer is the leading cause of cancer death among the African American population. Although many actionable biomarkers and targeted treatments are available to significantly prolong lung cancer survival, patients of non-European ancestry are less likely to undergo next-generation sequencing testing and received targeted

therapies than their white counterparts. Structural and environmental factors likely contribute to this disparity; however, there is evidence suggesting that somatic genomic differences also contribute – several lung cancer drivers and targetable biomarkers have been found to have different alteration frequencies between populations.

However, African American patients are severely underrepresented in research studies and clinical trials, and therefore, the genomic landscape of lung cancers with African ancestry remains poorly understood. Our preliminary analysis based on real-world, observational data of clinical tumor sequencing suggest that the

prevalence of targetable alterations in KRAS and ROS1 fusion are different in patients of African ancestry, independent to smoking status. Interestingly, ancestry may modify the effect of smoking exposure on the lung cancer genome – TP53 mutations were significantly enriched in smokers but not in never smokers of African

ancestry. Moreover, there is a significant enrichment of high tumor mutation burden in light smokers compared to their white counterparts, in line with previous studies showing higher risk of lung cancer in smokers of African ancestry, suggesting different genomic mechanism as the resultant influence of ancestry-smoking interaction.

In this project, we will take three parallel approaches to comprehensively characterize lung cancer genomes in patients of African ancestry. In the first aim, we will identify and validate more genomic differences between patients with African and European ancestry, with a focus on drug targetable and clinically relevant biomarkers.

We will also locus-specific germline ancestry to identify the genetic and non-genetic contributors to somatic differences. In the second aim, we will focus on investigating the interaction between African ancestry and smoking exposure on somatic alterations and develop a computational pipeline leveraging off-target reads from

panel sequencing to study genome-wide germline influence. In the third aim, we will develop a deep learning model to allow us further to understand the generalizability and ancestry-specificity of immunotherapy outcome prediction. Finally, we will move beyond panel sequencing to identify new markers for patients of African

ancestry, using whole-genome sequencing coupled with RNA-sequencing. The proposed study will broaden our knowledge about the complex relationship between genetic ancestry, environmental exposure and ancestry-environment interaction contributing to genomic differences. The findings will reveal germline factor associated with somatic phenotype, improve immunotherapy outcome prediction, and lead

to future study and discovery of new treatment options for African American patients.